What I Learned This Week

Machine learning faces a major roadblock on its path to widespread disruption: algorithmic accountability.

When the fake news firestorm ignited after the election, Facebook CEO Mark Zuckerberg called the idea that his platform’s algorithms influenced the outcome “crazy.” Then in April, he backtracked, telling The New York Times: “It’s something I think we’re still trying to figure out.” Regardless of the truth of the matter, the implication of Zuckerberg’s flip-flop is significant: the author of one of the world’s most-influential algorithms either won’t publicly acknowledge or doesn’t understand its influence. Today, algorithms guide everything from information consumption to the decisions investors, corporate leaders, and politicians make. Yet, scrutiny has been limited, or as Cathy O’Neil—for…

Want to read more?

Subscribe Today

Already a subscriber? Login here.